A5058720464- COVID-19, Geopolitics, Technology, Migration
- Regional Socio-Economic Development Trends
- Impact of AI and Big Data on Business and Society
- Sparse and Compressive Sensing Techniques
- Human Pose and Action Recognition
- Energy Efficient Wireless Sensor Networks
- Cooperative Communication and Network Coding
- Multimodal Machine Learning Applications
- Indoor and Outdoor Localization Technologies
- COVID-19 epidemiological studies
- Distributed Control Multi-Agent Systems
- Advanced MIMO Systems Optimization
- Target Tracking and Data Fusion in Sensor Networks
- Advanced Wireless Communication Techniques
- Hand Gesture Recognition Systems
- Distributed Sensor Networks and Detection Algorithms
- Data-Driven Disease Surveillance
- Video Analysis and Summarization
- Blind Source Separation Techniques
- Human Motion and Animation
- Wireless Communication Security Techniques
- Underwater Vehicles and Communication Systems
- Autism Spectrum Disorder Research
- Photoacoustic and Ultrasonic Imaging
- Millimeter-Wave Propagation and Modeling
Swiss Data Science Center
2020-2025
Paul Scherrer Institute
2022-2025
École Polytechnique Fédérale de Lausanne
2013-2023
ETH Zurich
2021-2022
Johns Hopkins University
2011-2020
Institute of Electrical and Electronics Engineers
2018
Signal Processing (United States)
2018
École Normale Supérieure - PSL
2014
Universidad Politécnica de Madrid
2007-2012
European Telecommunications Standards Institute
2008
<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Objective</i> : State-of-the-art techniques for surgical data analysis report promising results automated skill assessment and action recognition. The contributions of many these techniques, however, are limited to study-specific validation metrics, making progress across the field extremely challenging. xmlns:xlink="http://www.w3.org/1999/xlink">Methods</i> In this paper, we address two major...
Short-term forecasts of infectious disease burden can contribute to situational awareness and aid capacity planning. Based on best practice in other fields recent insights epidemiology, one maximise the predictive performance such if multiple models are combined into an ensemble. Here, we report ensembles predicting COVID-19 cases deaths across Europe between 08 March 2021 07 2022.
The characteristics of the power-line communication (PLC) channel are difficult to model due heterogeneity networks and lack common wiring practices. To obtain full variability PLC channel, random generators great importance for design testing algorithms. In this respect, we propose a generator that is based on top-down approach. Basically, describe multipath propagation coupling effects with an analytical model. We introduce into restricted set parameters and, finally, fit measured...
Energy efficiency is a major design issue in the context of Wireless Sensor Networks (WSN). If acquired data to be sent far-away base station, collaborative beamforming performed by sensors may help distribute communication load among nodes and reduce fast battery depletion. However, techniques are far from optimality many cases we might wasting more power than required. We consider energy applications. Using convex optimization framework, propose virtual beamformer that maximizes network...
Abstract. Emissions of harmful substances into the atmosphere are a serious environmental concern. In order to understand and predict their effects, it is necessary estimate exact quantity timing emissions from sensor measurements taken at different locations. There number methods for solving this problem. However, these existing assume Gaussian additive errors, making them extremely sensitive outlier measurements. We first show that errors in real-world measurement data sets come...
Since the beginning of COVID-19 pandemic, many dashboards have emerged as useful tools to monitor its evolution, inform public, and assist governments in decision-making. Here, we present a globally applicable method, integrated daily updated dashboard that provides an estimate trend evolution number cases deaths from reported data more than 200 countries territories, well 7-d forecasts. One significant difficulties managing quickly propagating epidemic is details dynamic needed forecast are...
Recent work on action recognition leverages 3D features and textual information to achieve state-of-the-art performance. However, most of the current few-shot methods still rely 2D frame-level representations, often require additional components model temporal relations, employ complex distance functions accurate alignment these representations. In addition, existing struggle effectively integrate semantics, some resorting concatenation or addition visual features, using text merely as an...
It often happens that we are interested in reconstructing an unknown signal from partial measurements. Also, it is typically assumed the location (temporal or spatial) of each sample known and only distortion present observations due to additive measurement noise. However, there some applications where such information lost. In this paper, consider situation which order noisy samples, taken a linear system, missing. Previous work on topic has considered noiseless case exhaustive search...
We propose a random channel generator for in-home power line communications (PLC). follow statistical top-down approach and we model the multipath propagation effects of PLC in frequency domain. Then, introduce variability into model, i.e., let parameters associated to reflections be random, according certain statistics. Finally, fit experimental data. target average path loss root-mean-square (RMS) delay spread measured channels. According this methodology, show that randomly generated...
Energy efficiency, scalability and robustness are key features of Ad-hoc Wireless Sensor Networks the use decentralized algorithms is practical importance in such scenarios. A method for node localization proposed by solving a nonlinear least-squares problem distributed fashion. For that purpose we propose Gauss-Newton algorithm with embedded consensus requires only local communication converges to centralized version.
We address the problem of sampling and reconstruction sparse signals with finite rate innovation. derive general conditions under which perfect is possible for kernels satisfying Strang-Fix conditions. Previous results on subject consider two particular cases; when kernel able to reproduce (complex) exponentials, or it has polynomial reproduction property. In this paper, we extend such analysis case where both properties could be found in show that former situations can regarded as special...
Abstract Background Short-term forecasts of infectious disease burden can contribute to situational awareness and aid capacity planning. Based on best practice in other fields recent insights epidemiology, one maximise the predictive performance such if multiple models are combined into an ensemble. Here we report ensembles predicting COVID-19 cases deaths across Europe between 08 March 2021 07 2022. Methods We used open-source tools develop a public European Forecast Hub. invited groups...
Abstract Hybrid Pixel Detectors (HPDs) are highly suitable in diffraction-based electron microscopy due to their high frame rates (> 1 kHz), dynamic range, and good radiation hardness. However, use imaging applications has been limited by relatively large pixel size (≥ 55 μm) high-energy (>80 keV) electrons scattering over multiple pixels the sensor layer. To realize full potential of fast, radiation-hard HPDs across modalities, we developed deep learning techniques precisely localize...
Mismatches between samples and their respective channel or target commonly arise in several real-world applications. For instance, whole-brain calcium imaging of freely moving organisms, multiple-target tracking multi-person contactless vital sign monitoring may be severely affected by mismatched sample-channel assignments. To address this issue systematically, we frame it as a signal reconstruction problem where correspondences channels are lost. Assuming sensing matrix for the signals,...
Recently, a new ICT paradigm emerged, which considers Multiple Devices that cooperate in Tasks (MDMT). Under this paradigm, cooperation among the nodes can be beneficial when subsets of share common interests or observations. For to successful, it is thus necessary account for decentralized labelling scheme allows uniquely identify every object interest. Such not only ensures proper data exchange but also formation interest-specific clusters and hence, might from communications cost...
Positioning in Wireless Sensor Networks is a key feature many applications. Finding efficient algorithms to perform this task of practical importance systems where limitations on the computational power and battery life are major issue. Forming coalitions within set visible nodes target can help reduce communication costs. We then formulate problem as coalitional game cooperation does not come for free.
Wireless sensor networks are posed as the new communication paradigm where use of small, low-complexity, and low-power devices is preferred over costly centralized systems. The spectra potential applications very wide, ranging from monitoring, surveillance, localization, among others. Localization a key application in simple, efficient, distributed algorithms paramount practical importance. Combining convex optimization tools with consensus we propose localization algorithm for scenarios...
Traditional sampling results assume that the sample locations are known. Motivated by simultaneous localization and mapping (SLAM) structure from motion (SfM), we investigate at unknown locations. Without further constraints, problem is often hopeless. For example, recently showed that, for polynomial bandlimited signals, it possible to find two arbitrarily far each other, fit measurements. However, also this can be overcome adding constraints positions. In paper, show these lead a uniform...